Rolling bearing fault diagnosis under fluctuant conditions based on compressed sensing

Abstract

Bearings are widely used in industries and construction for shaft supporting or seismic isolation. In recent years, their fault diagnosis, especially under variable, or fluctuant conditions, has received increasing attention. Sufficient monitoring data are usually required for bearing diagnosis. However, sufficient data cannot be guaranteed in some engineering cases with limitations of the transmission channel bandwidth or onboard/onsite computational capabilities. Fortunately, the emerging compressed sensing technique, which provides an effective solution to data compression and processing, has the ability to transform traditional monitoring data to the compressed information domain for a highly effective diagnosis under fluctuant conditions. This study proposes a bearing fault diagnosis method under fluctuant conditions based on compressed sensing theory. First, a random matrix is constructed as the measurement matrix and is employed to compress the original signal into the compressed information domain. Then, reconstruction-evaluation based fault diagnosis method is conducted with compressed signals. Moreover, the compressed signals used for fault diagnosis are reconstructed on the remote side. The experimental results provide evidence that the proposed method can effectively reduce the data volume required for bearing diagnosis and maintain an accuracy similar to current approaches, and the reconstructed signals can be used for other fault diagnosis methods.

Gear pitting severity level identification using binary segmentation methodology

“…The methodology presented in the current work can also be very helpful in identifying the change of states in other domains as well. For example, rolling bearings are used in seismic isolation devices to protect structural systems from earthquakes . For bearings fault diagnosis, the health indicator extracted using the vibration data collected on the bearing housing can be used to monitor the damage on bearing outer race and inner race.…”

Vibration health monitoring of rolling bearings under variable speed conditions by novel demodulation technique

“…Rolling bearing fault is one of the major reasons for a shutdown of machines. For reducing economic losses and avoiding accident, vibration‐based rolling bearing fault diagnosis has become an important topic in both academic and industrial areas for several decades 7–10 …”

Application of multivariate signal analysis in vibration‐based condition monitoring of wind turbine gearbox

“…Recently, Oliveira et al 2 proposed a new method based on the identification of the vibration modes of the modal parameters along 1 year monitoring and then utilized to investigate the study of the effects of natural frequencies on the operational and environmental condition. Yuan et al 3 studied a fault diagnosis of bearing under fluctuant conditions based on compressed sensing theory. Multi‐fault of wind turbine gearbox has been detected using complex wavelet transform 4 .…”